Can the Apple M4 (16 GB) run WeMM-Video-2B?
WeMM-Video-2B·3.3B·Apple M4 (16 GB)·16 GB unified memory
Yes. The Apple M4 (16 GB) runs WeMM-Video-2B fully in unified memory at FP16 / BF16 (about 7.3 GB of 11 GB), at roughly 10 tokens/s.
- best quant
- FP16 / BF16
- needs
- 7.3 GB
- usable
- 11 GB
- est. speed
- ~10 tok/s
Estimates for an 8K context; offload assumes 32 GB of system RAM.Try other contexts on the WeMM-Video-2B page →
Every quantisation of WeMM-Video-2B on the Apple M4 (16 GB)
| Quant | Memory | Fit | Speed |
|---|---|---|---|
| Q2_K | 1.9 GB | fits | ~61 tok/s |
| Q3_K_M | 2.4 GB | fits | ~41 tok/s |
| Q4_K_M | 2.8 GB | fits | ~33 tok/s |
| Q5_K_M | 3.2 GB | fits | ~28 tok/s |
| Q6_K | 3.5 GB | fits | ~24 tok/s |
| Q8_0 | 4.3 GB | fits | ~19 tok/s |
| FP16 / BF16 pick | 7.3 GB | fits | ~10 tok/s |
Speed is a bandwidth-based estimate for generating text (120 GB/s); prompt processing and long contexts are slower.
Measured speed on the Apple M4 (16 GB)
No community reports for this pair yet.
Report your own speed →Vision models that fit the Apple M4 (16 GB) comfortably
All that fit →| Model | Params | Best quant | Memory | Speed |
|---|---|---|---|---|
| 3.8B | Q8_0 | ≈4.0 GBFits in memory | ~17 tok/s | |
| 4.3B | Q8_0 | ≈5.2 GBFits in memory | ~14 tok/s | |
| 4.7B | Q8_0 | ≈5.9 GBFits in memory | ~13 tok/s | |
| 4.4B | Q8_0 | ≈5.8 GBFits in memory | ~14 tok/s | |
| 2.3B | Q8_0 | ≈2.8 GBFits in memory | ~27 tok/s | |
| 3.8B | Q8_0 | ≈6.9 GBFits in memory | ~16 tok/s |
Other GPUs that run WeMM-Video-2B
Frequently asked questions
- Can the Apple M4 (16 GB) run WeMM-Video-2B?
- Yes. The Apple M4 (16 GB) runs WeMM-Video-2B fully in unified memory at FP16 / BF16 (about 7.3 GB of 11 GB), at roughly 10 tokens/s.
- How fast does WeMM-Video-2B run on the Apple M4 (16 GB)?
- An estimate from memory bandwidth gives about 10 tokens/s at FP16 / BF16.
- Which quantisation of WeMM-Video-2B should I use on the Apple M4 (16 GB)?
- FP16 / BF16: the largest one that fits. Expected quality: lossless compared with the full-precision weights.
Memory and speed are estimates from the model's size and the card's memory bandwidth; real numbers depend on the runtime, drivers and context length. Community figures are medians of user reports.